PROJECT ABSTRACT Immune checkpoint blockade has revolutionized the treatment of cancers and demonstrated sustained and rapid efficacy in a number of patient populations. However, treatment outcomes can be variable across individual cancer patients or specific tumor types. For example, most patients with colorectal cancer (CRC) exhibit reduced efficacy to single agent checkpoint inhibitors. While this is partially accounted for by segregating patients based upon mismatch repair status and tumor mutation load, the mechanistic understanding of why most CRC patients do not respond to checkpoint blockade remains unclear. An additional explanation for this variability stems from recent seminal studies demonstrating a significant role for normally beneficial microbes, termed the microbiota, in controlling the efficacy of cancer checkpoint blockade in humans and mice. Despite these advances, the mechanisms by which host-microbiota interactions modulate anti-tumor immunity in the context of checkpoint blockade remains unclear, and the potential relevance of these pathways to specific tumor types has not been examined. The fundamental goals of this high- risk/high-reward proposal are to test a novel hypothesis that selective host-microbiota interactions critically promote resistance to checkpoint inhibitors in CRC, and that targeting this pathway will boost therapeutic responsiveness. We will comprehensively test this hypothesis using innovative patient samples and powerful mouse models. Results from these studies could lead to the development of novel combinatorial therapeutic approaches to boost the efficacy of cancer checkpoint inhibitors in multiple patient populations by targeting host-microbiota interactions.